English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Deep neural networks (DNN) have been successfully applied to musicclassification including music tagging. However, there are several openquestions regarding the training, evaluation, and analysis of DNNs. In thisarticle, we investigate specific aspects of neural networks, the effects ofnoisy labels, to deepen our understanding of their properties. We analyse and(re-)validate a large music tagging dataset to investigate the reliability oftraining and evaluation. Using a trained network, we compute label vectorsimilarities which is compared to groundtruth similarity.  The results highlight several important aspects of music tagging and neuralnetworks. We show that networks can be effective despite relatively large errorrates in groundtruth datasets, while conjecturing that label noise can be thecause of varying tag-wise performance differences. Lastly, the analysis of ourtrained network provides valuable insight into the relationships between musictags. These results highlight the benefit of using data-driven methods toaddress automatic music tagging.

The Effects of Noisy Labels on Deep Convolutional Neural Networks for Music Tagging